How to Write a Video Script with Midjourney

Midjourney is primarily an AI image generation tool, but it can become a powerful ally in writing a video script. By using Midjourney to create visual storyboards, scene concepts, and moods, you can structure your script around concrete images rather than abstract descriptions. This visual approach facilitates planning each sequence, helps communicate your vision to a production team, and significantly accelerates the creative process. In this tutorial, you will learn how to combine the visual power of Midjourney with classic scriptwriting techniques to produce a complete, illustrated document ready for shooting. Whether you are preparing a YouTube video, a short film, or an advertisement, this method will transform the way you script.

Prerequisites

1.An active Midjourney account with an active subscription
2.Access to Discord or the Midjourney web interface
3.A general idea of the subject or concept of your video
4.A text document (Google Docs, Notion, or Word) to write the script
5.Basic knowledge of narrative structure (introduction, development, conclusion)

Steps

Define the concept and narrative structure of the video

Before touching Midjourney, lay the foundations of your script. Identify the main message, target audience, and desired tone (educational, emotional, humorous). Divide your video into acts or sections: hook, introduction of the subject, development with 2 to 3 key points, climax or demonstration, and conclusion with a call to action. For each section, note the visual mood you imagine. This preparatory work will guide your Midjourney prompts and ensure the coherence of the final script.

Tip: Use the '5-second hook' method: your first scene must captivate immediately. Write the hook last, once the rest of the script is finished.

Generate a visual storyboard with Midjourney

For each section of your script, create an image that represents the key scene. Use Midjourney to generate visuals matching the mood, setting, and emotion of each sequence. Precisely describe the framing, lighting, and desired style in your prompts. Generate 2 to 4 variations per scene to explore different creative directions. Organize the selected images in chronological order of the script to form your storyboard.

/imagine a cinematic wide shot of [YOUR_SCENE], [VISUAL_STYLE], dramatic lighting, film still, 16:9 aspect ratio --ar 16:9 --style raw

Tip: Add '--ar 16:9' to get a standard video format. Use '--style raw' for more realistic and less stylized images, ideal for a production storyboard.

Write the script scene by scene based on the visuals

With your visual storyboard in front of you, write the script detailing each scene. For each generated image, write: the scene description (action, setting, camera movement), the voiceover text or dialogues, sound indications (music, sound effects), and the estimated duration. The Midjourney images serve as concrete visual references, making the writing more fluid and precise. Ensure the transitions between scenes are natural.

/imagine transition scene between [SCENE_A] and [SCENE_B], smooth cinematic movement, [MOOD], film photography --ar 16:9

Tip: Assign a duration to each scene from the drafting stage. A 3-minute video contains about 450 words of voiceover. Calibrate your text accordingly.

Refine the visual identity and script coherence

Use Midjourney's advanced features to harmonize the visual style across the entire storyboard. Regenerate images that stand out with similar parameters to the others. Use image references (--sref or --cref) to maintain consistency of style or character across scenes. In parallel, review the script to check tone consistency, narrative rhythm, and transition fluidity.

/imagine [SCENE_DESCRIPTION] --sref [REFERENCE_IMAGE_URL] --sv 3 --ar 16:9

Tip: The '--sref' parameter with a reference image helps maintain a consistent style. '--sv' controls the intensity of the style reference (1 to 4, 2-3 being the best compromise).

Assemble the final document and prepare for production

Compile all elements into a professional script document. Structure it as a two-column table: visuals on the left (inserted Midjourney images), audio on the right (voiceover, music, effects). Add metadata: title, total duration, format, target audience. Include a visual reference board in the appendix with all selected Midjourney images. This document will serve as a complete guide for shooting, editing, or creating the final video.

Tip: Export your Midjourney images in high resolution using the 'Upscale' function before integrating them into the final document. Name each file according to the scene number (scene-01.png, scene-02.png) to facilitate editing.

Common mistakes to avoid

✗Jumping into Midjourney without having defined the narrative structure beforehand, leading to inconsistent visuals and a disjointed script
✗Writing overly vague prompts like 'a beautiful scene' instead of precisely describing the framing, lighting, and desired mood
✗Neglecting the image format by forgetting the --ar 16:9 parameter, producing square visuals unusable for a video storyboard
✗Underestimating the duration of scenes and writing a script too long for the intended video format
✗Using Midjourney to generate text in images (titles, subtitles) when the tool handles typography poorly

FAQ

Can Midjourney write the video script text directly?

No, Midjourney is an image generator, not a text generator. It is used to create the visual storyboard and mood references that guide scriptwriting. For text (voiceover, dialogue), use a writing tool like ChatGPT or Claude in addition, then combine the text with Midjourney visuals.

How many Midjourney images should be generated for a 5-minute video script?

Count on average one key image every 15 to 30 seconds, i.e., between 10 and 20 images for a 5-minute script. Action scenes or quick transitions require more images, while a static dialogue scene may need only one.

Which Midjourney model is best for a video storyboard?

Use the latest model (v6 or higher) with the '--style raw' parameter for photorealistic images suitable for previewing. For animation or motion design videos, the standard mode (without --style raw) offers more stylized and creative results that are more suitable.

How to use this prompt

Copy the prompt with the button above.
Paste it into ChatGPT, Claude or your favorite AI assistant.
Replace the bracketed variables with your details, then refine the result.

About Prompt Guide

Prompt Guide is a free library of 4800+ ready-to-use prompts for ChatGPT, Claude and other AIs, with guides to learn prompting and tools to build and optimize your own prompts.

Prompt library Learn prompting Prompt builder Prompt optimizer

More tutorials

How to Write a Video Script with Mistral

Writing a compelling video script requires structure, creativity, and a good understanding of your audience. Mistral, the French AI assistant,

How to Write a Video Script with Perplexity

Writing a compelling video script requires thorough research, a clear structure, and a tone adapted to your audience. Perplexity, thanks to its real-time search and intelligent synthesis capabilities, becomes a powerful ally for this task.

How to Write a Video Script with Stable Diffusion

Stable Diffusion is a powerful AI image generation tool, but did you know it can also help you visually structure a video script? By combining Stable Diffusion's ability to create visual storyboards with a rigorous scripting methodology, you can pre-visualize every scene of your video before filming or editing. In this tutorial, we will guide you step-by-step to write a complete video script using Stable Diffusion as a creative assistant. You will learn how to define your concept, structure your sequences, generate reference visuals for each shot, and assemble everything into a professional production document. Whether you are a content creator, videographer, or marketer, this method will transform your creative process.

How to Write an Email with ChatGPT

Writing a professional or personal email can sometimes be time-consuming, especially when you're looking for the right tone and ideal wording. ChatGPT is a...

How to Write an Email with Claude

Writing a professional or personal email can sometimes be time-consuming, especially when you are looking for the right tone and ideal phrasing. Claude, the assistant...

How to Write an Email with DALL-E

DALL-E, developed by OpenAI, is primarily an AI image generation tool. Although it does not directly write text emails, it can significantly enrich your email communications by creating custom visuals, tailored illustrations, and striking graphic elements. In this tutorial, you will learn how to leverage DALL-E to design professional images that accompany your emails: banners, concept illustrations, promotional visuals, or header images. By combining DALL-E with your usual email client, you will transform ordinary emails into visually memorable communications that capture recipients' attention and enhance the impact of your message. Discover how to integrate generative AI into your email workflow.

Get new prompts every week

Join our newsletter.